Tips and Hints

People Getting Rich Online – Niche Research

How’s that list coming along? You wll recall in the first part of this series I covered getting a list together of ideas for your site or blog niche. Hopefully you have a good sized list of general categories. The next steps involve narrowing the keywords and then doing some supply and demand research.

There are many keyword tools out there. I like to use Overture’s because it’s free and easy. Keep in mind that the numbers from Overture are usually inflated — sometimes a little sometimes a lot. They fudge the numbers by counting like queries — fly and flies or affiliate and affiliates get counted as the same query. Remember — they make more money convincing their prospective advertisers that a search term is popular.

So plug in one of the general search terms from your list. I’m going to try dogs. Here are the results:

Searches done in February 2006
Count Search Term
 1104234  dog
 225786  dog breed
 183180 dog training basics
 152056  dog show
 139072  dog breeders
 98321  dog name
 97021  dog training
 91172  dog picture
 88480  dog for sale
 62657 dog grooming
 56907 dog gift

I edited a few like snoop dog out for expediency. I like dog breeds and dog training [basics]. Looks like people are searching for both quite a bit. Now that we have the demand, let’s go over to Google and check the supply.

Type in the exact search term, putting quotes around it so Google searches for exactly what we want. If you put in dog breeds, Google returns 5,59o,000 results. Look in the upper right-hand corner for this number.

Ever heard of Sumantra Roy? Probably not. He’s a man who specializes in SEO and particular keyword research. He has invented a formula called the Keyword Effectiveness Index or KEI to measure the potential of a keyword. I use a simpler version of it. Here it is:

KEI = Demand / Supply

That is – Demand divided by Supply. The higher the KEI the better. This gives you a simplified KEI that ranges from 1-10. People compare this to the Richter Scale, meaning that each successive number is exponentially better than the last number.

In our example above, the KEI for dog breed would be .04, which isn’t that great.

Dog Training Basics yields 45,600 results for Google. The KEI would be 4, which is terrific. Hmm…wait a minute. That concludes our lesson. Bye. [Sound of running down the hallway and door slamming]. A little nerd humor there.

Do a few more and see what you come up with. I did dog name and got a KEI of .10, which is lousy. I also tried dog gift, which got me a KEI of .07.

I usually pick up the highest KEI keywords and plug them back into the Overture tool and try it again.

Incidentally, Wordtracker does this all for you. Try it for free and see how easy it is. Just keep in mind that they use different sources for both supply and demand than my example above, but that shouldn’t make too much of a difference.

So now we should have a much smaller list that has some good prospects on it. In the next post we’ll get into the real research.

8 Responses to “People Getting Rich Online – Niche Research”

  1. Steve M. Says:

    I’ve also found it useful to go to some of the “live keyword” sites like Infotiger:

    to see different words and types of phrases that people use. It’s completely untargeted, but sometimes if gives you an idea about a turn of phrase or word association that you may not have thought of otherwise., synonyms and thesauruses are good help too.

  2. Jon Morrow Says:

    Eh? Maybe I’m missing something here. Demand (225,000) X Demand (225,000) / Supply (5.5 million) = 9204. Where do you get .08?

  3. Matt DeAngelis Says:

    Good catch. Actually I read the paper on KEI late last night and noticed that Mr. Roy’s formula was different from the one I use, so I plugged his in. I think the results are confusing (especially to me), so I plugged back in the one I have always used. The KEIs range from 1-10 and they are logarithmic, meaning that each successive number from 1-10 is much stronger than the number before it.

    In my haste to make sure you were getting the most correct formula, I used a more confusing one. So now it’s back to where it should be.


  4. aaron wall Says:

    I think the results are confusing

    I think it is easier just to look through the search results, then grab domain registration date (and look at the site history in, look at their top backlinks, and then see if they have any .gov or .edu backlinks.

    You really only have to look at the top few, as that is who you are competing with. Results 20+ are going to be irrelevant.

    It was not a highly commercialized word, but I ranked a site for a single word which has 63 million results on Google in under a month (myriadsearch is ranking for myriad). Result counts don’t tell you much.

  5. Matt DeAngelis Says:

    Not quite as scientific, but it works. I’m not sure I would base a $10,000 site launch on the results, but I’m sure it would work on, as you said, a non-highly commercialized word.

    I think results tell you a lot. Even if you go to the highest abstract level, they suggest the popularity of the words that you put in.

    Thanks for the suggestion, and congratulations on your success.

  6. aaron wall Says:

    Hi Matt
    I think my explanation sounds sorta backwards based on me not being good at explaining it (I must have been tired when I wrote that because my segway from point a to point b was non existant).

    I believe that as categories get more commercial KEI gets even less relevant as a proxy for competition. Why? Because they are less natural information pools. The top results are going to be far more competitive than KEI would estimate, and in some markets, such as mesothelioma, there are lots of junky automated or low quality pages that inflate the number of pages without adding any real competition.

    Looking through the inbound linkage data and site history is far more scientific than looking at the estimated number of results listed on a search results page.

    BTW…liked your post with the mention that you should reply to emails if u leave your email address published…sorta like smacked me in the head with a nice reminder.

  7. Matt DeAngelis Says:


    Not at all. I thought your comment was cogent and on-the-money, and I think you’re right this time too. The Google BigDaddy update is supposed to take care of some of this, but I think finding a lot of automated or low-quality pages in itself is an indicator of demand.

    The next part of my series is actually going to examine the type of research you’re talking about.

  8. » People Getting Rich Online - Finding Your Niche » -- Money-making insights for YOUR site Says:

    [...] People Getting Rich Online – Niche Research [...]